This notebook is an exercise in the Geospatial Analysis course. You can reference the tutorial at this link.

Introduction

You are an urban safety planner in Japan, and you are analyzing which areas of Japan need extra earthquake reinforcement. Which areas are both high in population density and prone to earthquakes?

Before you get started, run the code cell below to set everything up.

import pandas as pd
import geopandas as gpd

import folium
from folium import Choropleth, Circle, Marker
from folium.plugins import HeatMap

We define a function embed_map() for displaying interactive maps. It accepts two arguments: the variable containing the map, and the name of the HTML file where the map will be saved.

This function ensures that the maps are visible in all web browsers.

Exercises

1) Do earthquakes coincide with plate boundaries?

Run the code cell below to create a DataFrame plate_boundaries that shows global plate boundaries. The “coordinates” column is a list of (latitude, longitude) locations along the boundaries.

plate_boundaries = gpd.read_file(r"C:\Users\LG PC\Desktop\data_mining\archive\Plate_Boundaries\Plate_Boundaries\Plate_Boundaries.shp")
plate_boundaries['coordinates'] = plate_boundaries.apply(lambda x: [(b,a) for (a,b) in list(x.geometry.coords)], axis='columns')
plate_boundaries.drop('geometry', axis=1, inplace=True)

plate_boundaries.head()

	HAZ_PLATES	HAZ_PLAT_1	HAZ_PLAT_2	Shape_Leng	coordinates
0	TRENCH	SERAM TROUGH (ACTIVE)	6722	5.843467	[(-5.444200361999947, 133.6808931800001), (-5….
1	TRENCH	WETAR THRUST	6722	1.829013	[(-7.760600482999962, 125.47879802900002), (-7…
2	TRENCH	TRENCH WEST OF LUZON (MANILA TRENCH) NORTHERN …	6621	6.743604	[(19.817899819000047, 120.09999798800004), (19…
3	TRENCH	BONIN TRENCH	9821	8.329381	[(26.175899215000072, 143.20620700100005), (26…
4	TRENCH	NEW GUINEA TRENCH	8001	11.998145	[(0.41880004000006466, 132.8273013480001), (0….

Next, run the code cell below without changes to load the historical earthquake data into a DataFrame earthquakes.

# Load the data and print the first 5 rows
earthquakes = pd.read_csv(r"C:\Users\LG PC\Desktop\data_mining\archive\earthquakes1970-2014.csv", parse_dates=["DateTime"])
earthquakes.head()

	DateTime	Latitude	Longitude	Depth	Magnitude	MagType	NbStations	Gap	Distance	RMS	Source	EventID
0	1970-01-04 17:00:40.200	24.139	102.503	31.0	7.5	Ms	90.0	NaN	NaN	0.0	NEI	1.970010e+09
1	1970-01-06 05:35:51.800	-9.628	151.458	8.0	6.2	Ms	85.0	NaN	NaN	0.0	NEI	1.970011e+09
2	1970-01-08 17:12:39.100	-34.741	178.568	179.0	6.1	Mb	59.0	NaN	NaN	0.0	NEI	1.970011e+09
3	1970-01-10 12:07:08.600	6.825	126.737	73.0	6.1	Mb	91.0	NaN	NaN	0.0	NEI	1.970011e+09
4	1970-01-16 08:05:39.000	60.280	-152.660	85.0	6.0	ML	0.0	NaN	NaN	NaN	AK	NaN

The code cell below visualizes the plate boundaries on a map. Use all of the earthquake data to add a heatmap to the same map, to determine whether earthquakes coincide with plate boundaries.

# Create a base map with plate boundaries
m_1 = folium.Map(location=[35,136], tiles='cartodbpositron', zoom_start=5)
for i in range(len(plate_boundaries)):
    folium.PolyLine(locations=plate_boundaries.coordinates.iloc[i], weight=2, color='black').add_to(m_1)

# Your code here: Add a heatmap to the map
HeatMap(data=earthquakes[['Latitude', 'Longitude']], radius=10).add_to(m_1)

m_1

Make this Notebook Trusted to load map: File -> Trust Notebook

So, given the map above, do earthquakes coincide with plate boundaries?

2) Is there a relationship between earthquake depth and proximity to a plate boundary in Japan?

You recently read that the depth of earthquakes tells us important information about the structure of the earth. You’re interested to see if there are any intereresting global patterns, and you’d also like to understand how depth varies in Japan.

# Create a base map with plate boundaries
m_2 = folium.Map(location=[35,136], tiles='cartodbpositron', zoom_start=5)
for i in range(len(plate_boundaries)):
    folium.PolyLine(locations=plate_boundaries.coordinates.iloc[i], weight=2, color='black').add_to(m_2)
    
# Your code here: Add a map to visualize earthquake depth
def color_producer(val):
    if val <= 100:
        return 'forestgreen'
    else:
        return 'darkred'

# Add a bubble map to the base map
for i in range(0,len(earthquakes)):
    Circle(
        location=[earthquakes.iloc[i]['Latitude'], earthquakes.iloc[i]['Longitude']],
        radius=20,
        color=color_producer(earthquakes.iloc[i]['Depth'])).add_to(m_2)

# Display the map
m_2

Make this Notebook Trusted to load map: File -> Trust Notebook

Can you detect a relationship between proximity to a plate boundary and earthquake depth? Does this pattern hold globally? In Japan?

3) Which prefectures have high population density?

Run the next code cell (without changes) to create a GeoDataFrame prefectures that contains the geographical boundaries of Japanese prefectures.

# GeoDataFrame with prefecture boundaries
prefectures = gpd.read_file(r"C:\Users\LG PC\Desktop\data_mining\archive\japan-prefecture-boundaries\japan-prefecture-boundaries\japan-prefecture-boundaries.shp")
prefectures.set_index('prefecture', inplace=True)
prefectures.head()

	geometry
prefecture
Aichi	MULTIPOLYGON (((137.09523 34.65330, 137.09546 …
Akita	MULTIPOLYGON (((139.55725 39.20330, 139.55765 …
Aomori	MULTIPOLYGON (((141.39860 40.92472, 141.39806 …
Chiba	MULTIPOLYGON (((139.82488 34.98967, 139.82434 …
Ehime	MULTIPOLYGON (((132.55859 32.91224, 132.55904 …

The next code cell creates a DataFrame stats containing the population, area (in square kilometers), and population density (per square kilometer) for each Japanese prefecture. Run the code cell without changes.

# DataFrame containing population of each prefecture
population = pd.read_csv(r"C:\Users\LG PC\Desktop\data_mining\archive\japan-prefecture-population.csv")
population.set_index('prefecture', inplace=True)

# Calculate area (in square kilometers) of each prefecture
area_sqkm = pd.Series(prefectures.geometry.to_crs(epsg=32654).area / 10**6, name='area_sqkm')
stats = population.join(area_sqkm)

# Add density (per square kilometer) of each prefecture
stats['density'] = stats["population"] / stats["area_sqkm"]
stats.head()

	population	area_sqkm	density
prefecture
Tokyo	12868000	1800.614782	7146.448049
Kanagawa	8943000	2383.038975	3752.771186
Osaka	8801000	1923.151529	4576.342460
Aichi	7418000	5164.400005	1436.372085
Saitama	7130000	3794.036890	1879.264806

Use the next code cell to create a choropleth map to visualize population density.

# Create a base map
m_3 = folium.Map(location=[35,136], tiles='cartodbpositron', zoom_start=5)

# Your code here: create a choropleth map to visualize population density
Choropleth(geo_data=prefectures['geometry'],
           data=stats['density'], 
           key_on="feature.id", 
           fill_color='YlGnBu'
          ).add_to(m_3)

# Display the map
m_3

Make this Notebook Trusted to load map: File -> Trust Notebook

Which three prefectures have relatively higher density than the others? Are they spread throughout the country, or all located in roughly the same geographical region? (If you’re unfamiliar with Japanese geography, you might find this map useful to answer the questions.)

4) Which high-density prefecture is prone to high-magnitude earthquakes?

Create a map to suggest one prefecture that might benefit from earthquake reinforcement. Your map should visualize both density and earthquake magnitude.

# Create a base map
m_4 = folium.Map(location=[35,136], tiles='cartodbpositron', zoom_start=5)

# Your code here: create a map

Which prefecture do you recommend for extra earthquake reinforcement?

Keep going

Learn how to convert names of places to geographic coordinates with geocoding. You’ll also explore special ways to join information from multiple GeoDataFrames.

Have questions or comments? Visit the course discussion forum to chat with other learners.